Make sense of AI agents
Understand how agents work, give one a useful job, and test its boundaries. Three lessons built around a single practical example.
Start with a small job. Learn how agents work, set up OpenClaw, and follow practical exercises with results you can check.
From a fresh installation to an agent you can message. Install OpenClaw, connect Telegram, and learn what its permissions actually control.
Understand how agents work, give one a useful job, and test its boundaries. Three lessons built around a single practical example.
From a fresh installation to an agent you can message. Install OpenClaw, connect Telegram, and learn what its permissions actually control.
Install Emil Kowalski's skills, Impeccable and Taste, then connect the Figma and Playwright MCPs so Claude Code, Codex or Cursor designs, builds, tests and screenshots its own work.
Install TradingAgents, pick a model and a ticker, and read how its analyst, researcher, trader and risk agents reach a decision. Open source and free to run.
Install OmniRoute, connect free providers, and use them straight from Claude Code and Codex, with automatic fallback and token savings.
Understand the decision loop, choose a concrete job, and define where a person stays in control.
Choose a small tool set, decide what to remember, and make unfinished work easy for a person to pick up.
Write a delivery-reply brief and test it with three examples. A practical exercise before connecting real tools.
Ideas for work, everyday life, and learning—with a clear output to review. Choose one small job and make it useful.
Set up your own AI agent, connect a provider, and get your first reply. A practical guide from installation to a working chat.
Set up a Telegram bot, approve your account, and verify a reply on your phone. Includes paths for Discord and WhatsApp.
Inspect access, set a deliberate command policy, and check the result of a small task against its activity.